Loading ...
Sorry, an error occurred while loading the content.
 

Re: [wmlprogramming] Re: Cache control extensibility

Expand Messages
  • Luca Passani
    ... So, what drives you to go out of your way to defend transcoders and find any kind of corner cases to legitimize their existence? For the record, I still
    Message 1 of 14 , Nov 29, 2008
      Tom Hume wrote:
      > Yawn. You've used the "snide inference that I must be in the pay of
      > the transcoding industry" line before, Luca - on this list back in
      > August. I'm not. Then I answered you thus:
      >
      >
      >>> Never worked for a transcoder vendor, never had one as a customer, no
      >>> other affiliation.
      >>>
      >>> My company has worked for O2, Orange, Vodafone, 3, and T-Mobile. So
      >>> certainly one of those counts as an operator who's installed an
      >>> abusive transcoder (tho I'd point out I'm on record as saying I
      >>> though
      >>> the Voda/Novarra installation was unhelpful and a Bad Thing).
      >>>

      So, what drives you to go out of your way to defend transcoders and find
      any kind of corner cases to legitimize their existence?

      For the record, I still think you are a collaborationist.

      >
      > What I'm saying in the article you link to is:
      >
      > 1. Transcoding is not in itself illegal, which is what the only legal
      > opinion we have confirms;
      >
      it is illegal when they transcode sites which rely on recognizing the UA
      string, because this behavior prevents them from serving the right content.

      > 2. It should be opt-in *only* (to avoid a repeat of Vodafone/Novarra);
      >

      real opt-in transcoders do not need CTG. They do what they want. If
      Novarra has spent resources to push CTG, it means that they need to use
      it for non-opt-in transcoders too.


      > 3. "no-transform" must be respected.
      >
      yes, but MUST is supposed to be a must for transcoders. Automagically,
      you and your fellow W3C bullshitters are turning it into an obligation
      for content owners.

      > @legalien also raised what I found the first compelling use case I've
      > seen for someone wanting a transcoded web experience over a made-for-
      > mobile, when he/she pointed out that web sites send to have more
      > content than made-for-mobile experiences, and that end-users might
      > want to access this content, instead of a stripped down mobile
      > version, on the move.
      >
      yes, and I have argued that users should have access to a transcoded
      experience if and only if the content owner has agreed (explicitly or
      implicitly) to let the user have it. This is an obvious consequence of a
      basic truth: the content owner is king (and the W3C counsel agrees with
      me about this).

      > Yes, Cache-control is the defence against transcoding. It's been in
      > HTTP for years (certainly predating the mobile transcoding industry),
      > and it's *the* way to say "don't transform my content". It *must* be
      > respected by transcoder vendors.
      >
      What about WML? W3C telling developers to use cache-control is like a
      doctor telling a patient that this or that medicine may help in some
      cases, but it will cause heart failure in some contexts. This doctor
      should be put away before s/he causes more damage.

      > It's not all that is needed - folks here and the CTG document itself
      > agree that we need more than no-transform, to allow signalling to
      > other parties in the request/response chain that some transformations
      > are acceptable and others not. But we don't have those now and it
      > would be kind of weird to put something which is of zero use today
      > into a "best practices" document.
      >
      Look. Stop bullshitting us. The correct way for transcoders to behave if
      they were honest (which they obviously aren't) would be to wait until a
      new standard exist and preserve mobile sites until that day. CTG helps
      the abusers perpetrate their abuse. This has been demonstrated beyond
      reasonable doubt.

      > I'm putting in effort because, like you, I care about this issue. Your
      > repeated assertions that W3C/CTG is engaging in some sort of
      > conspiracy or being used by the transcoding industry are false, and I
      > don't want to see comments to that effect go unchecked.
      >
      it's not false. Everything has happened exactly the way I had predicted.
      I predicted that CTG would be used by Novarra to justify abusive
      installations of their shitvision proxy, and this is exactly what happened.

      > Luca, when you say "Please do not start to quarrel with one another.
      > We are all on the same boat here.", I think you are spot on.
      Let me rephrase: "We all except Tom Hume are on the same boat here. Tom
      is on some boat of his own licking operator ass".

      > We are
      > all in the same boat, and we're all working towards the same goal.
      >
      You are pushing in the opposite direction. Again. Stop bullshitting me.

      Luca
    • Tom Hume
      ... I don t think I am defending them, actually. I can see a use case where I personally might want transcoding - it s limited, but it s there. I think they
      Message 2 of 14 , Nov 29, 2008
        On 29 Nov 2008, at 17:08, Luca Passani wrote:

        > So, what drives you to go out of your way to defend transcoders and
        > find
        > any kind of corner cases to legitimize their existence?
        >
        I don't think I am defending them, actually.

        I can see a use case where I personally might want transcoding - it's
        limited, but it's there. I think they should be regulated or self-
        regulate in some way to avoid future Novarra/Vodafone-style damage,
        and I think that by default users should always get the best mobile
        experience available (which won't be a transcoded web experience
        unless nothing else is available).
        > > What I'm saying in the article you link to is:
        > > 1. Transcoding is not in itself illegal, which is what the only
        > legal
        > > opinion we have confirms;
        > it is illegal when they transcode sites which rely on recognizing
        > the UA
        > string, because this behavior prevents them from serving the right
        > content.
        >
        It isn't illegal to change the UA string.

        If you believe that the transcoding industry is in and of itself
        illegal, there is no case for "responsible reformatting", a manifesto,
        or a CTG document; it should never occur, ever, and you should be
        lobbying to have Novarra, Openwave, Infogin, etc. sued.

        Is your position that transcoding is in and of itself wrong, or that
        transcoder deployments need to be responsible and regulated?
        > > 3. "no-transform" must be respected.
        > yes, but MUST is supposed to be a must for transcoders. Automagically,
        > you and your fellow W3C bullshitters are turning it into an obligation
        > for content owners.
        >
        Can you explain what you mean there? I'm having trouble divining sense
        from it.
        > yes, and I have argued that users should have access to a transcoded
        > experience if and only if the content owner has agreed (explicitly or
        > implicitly) to let the user have it. This is an obvious consequence
        > of a
        > basic truth: the content owner is king (and the W3C counsel agrees
        > with
        > me about this).
        >
        That's not *exactly* what Rigo said though, is it? He said that no-
        transform was the mechanism to say "don't transform my content", and
        that providing such a mechanism (which must be adhered to) allows
        content providers to assert their rights to avoid transformation. That
        sounds like sound advice to me: avoid transformation by using no-
        transform directives. If you don't, your content may be transformed.
        > > Yes, Cache-control is the defence against transcoding. It's been in
        > > HTTP for years (certainly predating the mobile transcoding
        > industry),
        > > and it's *the* way to say "don't transform my content". It *must* be
        > > respected by transcoder vendors.
        > What about WML?
        >
        WML shouldn't be transformed or transcoded IMHO. This is a point that
        Eduardo has raised recently on the public-bpwg-ct list.
        > it's not false. Everything has happened exactly the way I had
        > predicted.
        > I predicted that CTG would be used by Novarra to justify abusive
        > installations of their shitvision proxy, and this is exactly what
        > happened.
        >
        Verizon are not Novarra; it was Verizon that produced that document.
        This does, of course, raise the issue of where responsibility lies for
        misbehaving transcoder deployments; if it lies with operators (and I'd
        say it does, personally), then it's important to persuade them to do
        so. Perhaps you should consider putting effort into having operators
        sign the Manifesto, in that case?

        Tom

        P.S. Thanks for the use of the word "shitvision" - that's absolutely
        top swearing, and I plan to use it myself :)

        --
        Future Platforms Ltd
        e: Tom.Hume@...
        t: +44 (0) 1273 819038
        m: +44 (0) 7971 781422
        company: www.futureplatforms.com
        personal: tomhume.org
      • David Harper
        @Tom. Re: and that providing such a mechanism (which must be adhered to) allows content providers to assert their rights to avoid transformation. That sounds
        Message 3 of 14 , Nov 29, 2008
          @Tom.

          Re: 'and that providing such a mechanism (which must be adhered to)
          allows content providers to assert their rights to avoid
          transformation. That sounds like sound advice to me: avoid
          transformation by using no- transform directive"

          These mechanisms are not adhered to.

          Cheers,
          David Harper
          Founder & CEO, Winksite


          --- In wmlprogramming@yahoogroups.com, Tom Hume <Tom.Hume@...> wrote:
          >
          >
          > On 29 Nov 2008, at 17:08, Luca Passani wrote:
          >
          > > So, what drives you to go out of your way to defend transcoders and
          > > find
          > > any kind of corner cases to legitimize their existence?
          > >
          > I don't think I am defending them, actually.
          >
          > I can see a use case where I personally might want transcoding - it's
          > limited, but it's there. I think they should be regulated or self-
          > regulate in some way to avoid future Novarra/Vodafone-style damage,
          > and I think that by default users should always get the best mobile
          > experience available (which won't be a transcoded web experience
          > unless nothing else is available).
          > > > What I'm saying in the article you link to is:
          > > > 1. Transcoding is not in itself illegal, which is what the only
          > > legal
          > > > opinion we have confirms;
          > > it is illegal when they transcode sites which rely on recognizing
          > > the UA
          > > string, because this behavior prevents them from serving the right
          > > content.
          > >
          > It isn't illegal to change the UA string.
          >
          > If you believe that the transcoding industry is in and of itself
          > illegal, there is no case for "responsible reformatting", a manifesto,
          > or a CTG document; it should never occur, ever, and you should be
          > lobbying to have Novarra, Openwave, Infogin, etc. sued.
          >
          > Is your position that transcoding is in and of itself wrong, or that
          > transcoder deployments need to be responsible and regulated?
          > > > 3. "no-transform" must be respected.
          > > yes, but MUST is supposed to be a must for transcoders. Automagically,
          > > you and your fellow W3C bullshitters are turning it into an obligation
          > > for content owners.
          > >
          > Can you explain what you mean there? I'm having trouble divining sense
          > from it.
          > > yes, and I have argued that users should have access to a transcoded
          > > experience if and only if the content owner has agreed (explicitly or
          > > implicitly) to let the user have it. This is an obvious consequence
          > > of a
          > > basic truth: the content owner is king (and the W3C counsel agrees
          > > with
          > > me about this).
          > >
          > That's not *exactly* what Rigo said though, is it? He said that no-
          > transform was the mechanism to say "don't transform my content", and
          > that providing such a mechanism (which must be adhered to) allows
          > content providers to assert their rights to avoid transformation. That
          > sounds like sound advice to me: avoid transformation by using no-
          > transform directives. If you don't, your content may be transformed.
          > > > Yes, Cache-control is the defence against transcoding. It's been in
          > > > HTTP for years (certainly predating the mobile transcoding
          > > industry),
          > > > and it's *the* way to say "don't transform my content". It *must* be
          > > > respected by transcoder vendors.
          > > What about WML?
          > >
          > WML shouldn't be transformed or transcoded IMHO. This is a point that
          > Eduardo has raised recently on the public-bpwg-ct list.
          > > it's not false. Everything has happened exactly the way I had
          > > predicted.
          > > I predicted that CTG would be used by Novarra to justify abusive
          > > installations of their shitvision proxy, and this is exactly what
          > > happened.
          > >
          > Verizon are not Novarra; it was Verizon that produced that document.
          > This does, of course, raise the issue of where responsibility lies for
          > misbehaving transcoder deployments; if it lies with operators (and I'd
          > say it does, personally), then it's important to persuade them to do
          > so. Perhaps you should consider putting effort into having operators
          > sign the Manifesto, in that case?
          >
          > Tom
          >
          > P.S. Thanks for the use of the word "shitvision" - that's absolutely
          > top swearing, and I plan to use it myself :)
          >
          > --
          > Future Platforms Ltd
          > e: Tom.Hume@...
          > t: +44 (0) 1273 819038
          > m: +44 (0) 7971 781422
          > company: www.futureplatforms.com
          > personal: tomhume.org
          >
        • Tom Hume
          David Yes - and that s absolutely wrong. They should be. CTG, Manifesto, HTTP and legal advice all agree on this point. Tom ... -- Future Platforms Ltd e:
          Message 4 of 14 , Nov 29, 2008
            David

            Yes - and that's absolutely wrong. They should be. CTG, Manifesto,
            HTTP and legal advice all agree on this point.

            Tom

            On 29 Nov 2008, at 21:54, David Harper wrote:

            > @Tom.
            >
            > Re: 'and that providing such a mechanism (which must be adhered to)
            > allows content providers to assert their rights to avoid
            > transformation. That sounds like sound advice to me: avoid
            > transformation by using no- transform directive"
            >
            > These mechanisms are not adhered to.
            >
            > Cheers,
            > David Harper
            > Founder & CEO, Winksite
            >
            > --- In wmlprogramming@yahoogroups.com, Tom Hume <Tom.Hume@...> wrote:
            > >
            > >
            > > On 29 Nov 2008, at 17:08, Luca Passani wrote:
            > >
            > > > So, what drives you to go out of your way to defend transcoders
            > and
            > > > find
            > > > any kind of corner cases to legitimize their existence?
            > > >
            > > I don't think I am defending them, actually.
            > >
            > > I can see a use case where I personally might want transcoding -
            > it's
            > > limited, but it's there. I think they should be regulated or self-
            > > regulate in some way to avoid future Novarra/Vodafone-style damage,
            > > and I think that by default users should always get the best mobile
            > > experience available (which won't be a transcoded web experience
            > > unless nothing else is available).
            > > > > What I'm saying in the article you link to is:
            > > > > 1. Transcoding is not in itself illegal, which is what the only
            > > > legal
            > > > > opinion we have confirms;
            > > > it is illegal when they transcode sites which rely on recognizing
            > > > the UA
            > > > string, because this behavior prevents them from serving the right
            > > > content.
            > > >
            > > It isn't illegal to change the UA string.
            > >
            > > If you believe that the transcoding industry is in and of itself
            > > illegal, there is no case for "responsible reformatting", a
            > manifesto,
            > > or a CTG document; it should never occur, ever, and you should be
            > > lobbying to have Novarra, Openwave, Infogin, etc. sued.
            > >
            > > Is your position that transcoding is in and of itself wrong, or that
            > > transcoder deployments need to be responsible and regulated?
            > > > > 3. "no-transform" must be respected.
            > > > yes, but MUST is supposed to be a must for transcoders.
            > Automagically,
            > > > you and your fellow W3C bullshitters are turning it into an
            > obligation
            > > > for content owners.
            > > >
            > > Can you explain what you mean there? I'm having trouble divining
            > sense
            > > from it.
            > > > yes, and I have argued that users should have access to a
            > transcoded
            > > > experience if and only if the content owner has agreed
            > (explicitly or
            > > > implicitly) to let the user have it. This is an obvious
            > consequence
            > > > of a
            > > > basic truth: the content owner is king (and the W3C counsel agrees
            > > > with
            > > > me about this).
            > > >
            > > That's not *exactly* what Rigo said though, is it? He said that no-
            > > transform was the mechanism to say "don't transform my content", and
            > > that providing such a mechanism (which must be adhered to) allows
            > > content providers to assert their rights to avoid transformation.
            > That
            > > sounds like sound advice to me: avoid transformation by using no-
            > > transform directives. If you don't, your content may be transformed.
            > > > > Yes, Cache-control is the defence against transcoding. It's
            > been in
            > > > > HTTP for years (certainly predating the mobile transcoding
            > > > industry),
            > > > > and it's *the* way to say "don't transform my content". It
            > *must* be
            > > > > respected by transcoder vendors.
            > > > What about WML?
            > > >
            > > WML shouldn't be transformed or transcoded IMHO. This is a point
            > that
            > > Eduardo has raised recently on the public-bpwg-ct list.
            > > > it's not false. Everything has happened exactly the way I had
            > > > predicted.
            > > > I predicted that CTG would be used by Novarra to justify abusive
            > > > installations of their shitvision proxy, and this is exactly what
            > > > happened.
            > > >
            > > Verizon are not Novarra; it was Verizon that produced that document.
            > > This does, of course, raise the issue of where responsibility lies
            > for
            > > misbehaving transcoder deployments; if it lies with operators (and
            > I'd
            > > say it does, personally), then it's important to persuade them to do
            > > so. Perhaps you should consider putting effort into having operators
            > > sign the Manifesto, in that case?
            > >
            > > Tom
            > >
            > > P.S. Thanks for the use of the word "shitvision" - that's absolutely
            > > top swearing, and I plan to use it myself :)
            > >
            > > --
            > > Future Platforms Ltd
            > > e: Tom.Hume@...
            > > t: +44 (0) 1273 819038
            > > m: +44 (0) 7971 781422
            > > company: www.futureplatforms.com
            > > personal: tomhume.org
            > >
            >
            >
            >

            --
            Future Platforms Ltd
            e: Tom.Hume@...
            t: +44 (0) 1273 819038
            m: +44 (0) 7971 781422
            company: www.futureplatforms.com
            personal: tomhume.org





            [Non-text portions of this message have been removed]
          • Luca Passani
            ... I think the use cases where a transcoders can be used are much more limited than transcoder vendors would like us to believe. Transcoders have existed for
            Message 5 of 14 , Nov 29, 2008
              Tom Hume wrote:
              > If you believe that the transcoding industry is in and of itself
              > illegal, there is no case for "responsible reformatting", a manifesto,
              > or a CTG document; it should never occur, ever, and you should be
              > lobbying to have Novarra, Openwave, Infogin, etc. sued.
              >
              I think the use cases where a transcoders can be used are much more
              limited than transcoder vendors would like us to believe. Transcoders
              have existed for some time, but never before would they become as bold
              as being marketed as a replacement for the whole mobile web. The
              Manifesto is a compromise which at least managed to protect
              mobile-optimised sites.

              > Is your position that transcoding is in and of itself wrong, or that
              > transcoder deployments need to be responsible and regulated?
              >
              I repeat: there may be legitimate uses for transcoding. Lately they have
              been used abusively. The Manifesto offers decent protection. CTG is not
              a good enough answer for a variety of reason, first one of which is that
              it does not mandate UA preservation.

              >>> 3. "no-transform" must be respected.
              >>>
              >> yes, but MUST is supposed to be a must for transcoders. Automagically,
              >> you and your fellow W3C bullshitters are turning it into an obligation
              >> for content owners.
              >>
              >>
              > Can you explain what you mean there? I'm having trouble divining sense
              > from it.
              >

              Sure. When we introduced the "no-transform" rule in the Manifesto, it
              was meant to be extra-protection for developers. i.e. a further
              requirement for transcoders.
              With CTG and W3C, it has become mandatory for developers to add the
              header: as you and Rigo have expressed, if you do not add the
              no-transform header, your content is cannon fodder and transcoders can
              transform it.


              >> yes, and I have argued that users should have access to a transcoded
              >> experience if and only if the content owner has agreed (explicitly or
              >> implicitly) to let the user have it. This is an obvious consequence
              >> of a
              >> basic truth: the content owner is king (and the W3C counsel agrees
              >> with
              >> me about this).
              >>
              >>
              > That's not *exactly* what Rigo said though, is it? He said that no-
              > transform was the mechanism to say "don't transform my content", and
              > that providing such a mechanism (which must be adhered to) allows
              > content providers to assert their rights to avoid transformation. That
              > sounds like sound advice to me: avoid transformation by using no-
              > transform directives. If you don't, your content may be transformed.
              >
              yes. Rigo says that you can use no-transform. I strongly disagree. To
              me, this is not OK. no-transform introduces extra issues for developers
              and cannot be mandated the way CTG is implicitly doing.


              > WML shouldn't be transformed or transcoded IMHO. This is a point that
              > Eduardo has raised recently on the public-bpwg-ct list.
              >
              yes, but there is a little problem. The Note in 4.2.2:

              http://www.w3.org/TR/ct-guidelines/#sec-cache-control-no-transform

              "Including a |Cache-Control: no-transform| directive can disrupt the
              behavior of WAP/WML proxies, because it can inhibit such proxies from
              converting WML to WMLC."

              CTG is basically saying "we know that our recommendation may produce
              serious problems with your existing application, but now that we told
              you the problem is magically all yours to manage"


              > Verizon are not Novarra; it was Verizon that produced that document.
              >
              Don't be naive. Verizon knows nothing about CTG. If they wrote that
              stuff is because someone from Novarra has written it for them. And since
              very few also in Novarra know about CTG, the (co-)author is easily
              tracked: Sean Patterson.

              > This does, of course, raise the issue of where responsibility lies for
              > misbehaving transcoder deployments; if it lies with operators (and I'd
              > say it does, personally), then it's important to persuade them to do
              > so. Perhaps you should consider putting effort into having operators
              > sign the Manifesto, in that case?
              >
              The effort is under the way. There are a few large ones that are
              seriously considering. But I did not manage to get anyone high enough to
              take the responsibility (problem is, there is no incentive to them in
              signing)


              >
              > P.S. Thanks for the use of the word "shitvision" - that's absolutely
              > top swearing, and I plan to use it myself :)
              >
              not bad for someone who speaks english a second language, is it?

              Luca
            Your message has been successfully submitted and would be delivered to recipients shortly.